Hari User

At Brillio, I built a self-service Terraform and CloudFormation catalog that cut environment provisioning from 3 days to 45 minutes. I also reduced annual cloud spend by about 18% across an AWS footprint of more than $500K.

I own multi-cluster Amazon EKS platform work, sustaining 99.9% availability across more than 20 production microservices. At Brillio, I also improved observability and service mesh delivery, reducing MTTD from 15 to 2 minutes and MTTR from 45 to 12 minutes.

I built an MLflow pipeline with Amazon SageMaker and KServe that reduced model retraining and release turnaround from 2 weeks to 30 minutes. I also designed cross-region EKS disaster recovery, achieving a verified 15-minute RPO and 30-minute RTO for business-critical workloads.

At Accenture, I migrated a healthcare platform into 45 containerized microservices on Amazon EKS, serving more than 300 hospitals under SOC 2 and HIPAA requirements. I’ve also contributed to Falco runtime security rules and ArgoCD documentation, and published open-source Kubernetes and MLOps reference architectures.